摘要
在深入研究尺度小波变换和嵌入式小波零树压缩编码算法的基础上 ,探讨了一种改进的小波零树编码算法 .通过采用基于嵌入式零树预测扫描算法的改进小波变换编码方法 ,对小波分解的低频子带图像采用DPCM预测编码 ,高频子带图像采用了改进的取消重要系数排序的零树预测扫描算法进行压缩编码 ,使之既具有原零树算法良好的压缩品质 ,同时也简化了原压缩算法 ,提高了该算法的适应性 ,与 JPEG压缩相比 ,它消除了JPEG图像在高压缩比下的方块效应 。
We proposed the improved embedding zerotree wavelet algorithm after a thorough analysis of the decomposed wavelet coefficients and the original EZW algorithm. In the proposed algorithm, the low\|frequency sub\|image wavelet coefficients are processed using predictive coding, meanwhile, the high\|frequency ones are processed using the improved EZW algorithm, having the ordering operation of the significant coefficient canceled. The improved algorithm not only bears the effective compression performance of the original algorithm, but also simplifies the algorithm with the reduced processing time of image encoding and decoding and enhanced practical performance as well. The experimental results are appreciated.
出处
《武汉大学学报（理学版）》
CAS
CSCD
北大核心
2002年第1期111-115,共5页
Journal of Wuhan University:Natural Science Edition
基金
武汉市科委重点项目资助 ( 9910 0 5 115 )
